Autism is a severely debilitating developmental disorder with an onset early in life. Prominent among the children's early impairments are deficits in sharing experiences with others in joint attention. As the children get older, they also manifest characteristic cognitive deficits, especially in meta-representation and executive function. A primary aim of the proposed research is to elucidate how these different facets of the syndrome are related to one another, through studying a process on the interface between social and cognitive domains of psychological functioning - symbolic play. Although many children with autism are able to 'pretend' under structured conditions or when directly instructed to do so, their play lacks spontaneous movement, creativity, and imagination. The identification theory of autism holds that because of primary deficits in socio-emotional responsiveness, children with autism are seldom 'moved' to adopt the psychological perspectives of others. Our hypothesis is that this weaker capacity to 'identify with' other people may compromise the development of their abilities in executive function (flexible thinking and inhibition of actions) and meta-representation (understanding how people create new meanings and apply them to objects in play). A specific difficulty in adopting new perspectives or meanings through other people may have important consequences for the children's relative limitations in the initiation of differing 'takes' on the world, including toward the self. Thus, this project will study how children adopt and share new meanings in a social play context, and explore whether such socially grounded shifts mediate relations between identification and critical aspects of cognitive functioning. We shall recruit 30 young children with autism and 30 age- and language-matched children without autism who have developmental disorders (DD). We shall assess the children's capacity for meta-representation, executive function skills, and their propensity to 'identify with' (adopt the perspective of) others in imitation and communication. In addition, we will conduct an in-depth investigation of spontaneous meaning shifts (symbolic movement) during a semi-structured play task. We predict between-group differences in specific aspects of each domain. Further, we predict a) there will be group differences in symbolic movement between subsets of children group-matched for meta-representation abilities, b) symbolic movement will be associated with identification, even when meta-representation and executive function abilities are taken into account, c) symbolic movement in play will be more likely to occur during episodes of joint attention, and specifically those involving experience-sharing, and d) this form of symbolic movement - socially-derived meaning shifts - will mediate relations both between identification and meta-representation, and between identification and executive function skills. Taken together, this pattern of findings would suggest that creativity in play is a socially-grounded phenomenon. Such findings would have direct clinical significance for the focus and scope of social interventions to facilitate cognitive development among children with autism. [unreadable] [unreadable] [unreadable]